How do you track exercise calories?

I don't use MFP's exercise calorie estimates for many things.

When first using MFP, I tried to learn about various ways of estimating calories, then identified what I thought were the most accurate ways of estimating each type of exercise I did routinely.

Many people do well syncing a good fitness tracker to MFP, and letting MFP and the tracker sort out all the activity calories between them. I don't do that because I turned out to be statistically odd in terms of calorie needs: I have a fitness tracker that others here have reported as estimating pretty well for them, but it's off for me by 25-30%, which is hundreds of calories daily, compared to 10+ years of careful eating/weight logging results.

The only case where I routinely use MFP's calorie estimate is for standard reps/sets strength training.

For intentional walking for exercise, I use this calculator, setting the energy box on "net": https://exrx.net/Calculators/WalkRunMETs

For stationary biking, I use (hours x 3.6 x average watts).

For outdoor biking, on-water rowing, and machine rowing, I use Garmin's exercise calorie estimate for the workout.

In general, I pick the lowest of the plausible estimating methods for an activity, but I also think about the underlying estimating methods to consider whether they're a realistic way to estimating that particular thing. For example, heart rate is a terrible way to estimate strength training calories. METS are usually a bad way to estimate things that don't vary by body weight in terms of physics-type work. And so forth.

There's more to it than that, but that's basically what I do.

I suggest picking a conservative number for exercise for a couple of key reasons. First, it's very easy to over-estimate exercise calories burned. Second, even if you get an accurate number, that doesn't mean you should eat back 100% of those calories. For example, you might burn 500 more calories from exercise and then burn 250 fewer calories in the rest of the day from reduced NEAT.

So pick a conservative number, track your weight progress, and go from there. Remember that your daily calorie target from a calculator is an estimate itself, before adding in the other estimate of your exercise. Lots of room for errors and assumptions to creep in.

Basically, if my exercise is consistent and my calories in are X, what is my weight change? With that info, adjust over time as required.

Years of experience and logging food carefully. What's left from weightloss and calorie intake is exercise calories, basically. I find that my Garmin watch works quite well for my running exercise calories. Walking is way off, but I don't often do really long walks where it might count. I use a fixed amount for my current strength routine. Basically, if I lose faster then I can eat more, if I lose slower I overestimate. Note, this is over several weeks, not a few days.

I don't, thats just one more number that will never be accurate. If your exercise program is fairly consistent then just figure it into your TDEE.

Is your exercise program coupled with your overall activity at a certain weekly calorie amount achieving your desired results? Then don't overcomplicate things with more numbers.

Adjust as needed…..

That's a little generic. Maybe it's not, but Grammerly's AI detector says this looks like it could be written by AI. I think reality is a little more nuanced . . . at least that's been my experience.

For one, exercise calories don't have much to do with fitness progress, i.e., getting stronger muscularly, cardiovascularly, in mobility or the like. Yes, exercise calories have a lot to do with weight management goals, and body weight has some impact on fitness, but not as much as many people might imagine. (I'm saying that as someone who was once the semi-mythical pretty-fit obese person, athletically very active and with good fitness markers despite excess body fat.)

Fitness trackers don't necessarily provide "accurate estimates". It varies by individual and exercise type. Heart rate in particular is fraught, because most people don't know their actual maximimum heart rate. The devices generally use age estimates, and those age estimates are meaningfully inaccurate for a large minority of the population.

Many online exercise calorie calculators and gym machines include calories in their estimate that are already included in our base MFP calorie estimate - BMR/RMR calories, plus the calories we would've burned just doing daily life stuff if not exercising in that time slice. Some gym machines wildly over-estimate calorie burn - kind of useful for marketing, y'know? If we log those calories that are over-estimated, it will affect our weight loss rate.

There has been some pretty nuanced, personal-experience-based advice on this thread, from people who've tracked and succeeded with their goals while using tracking as one of their tools. When I was losing weight, even now in long term maintenance, that's the kind of advice that was most valuable to me. I couldn't get it anywhere else, except from social sources like this Community. This one in particular is IMO more helpful than some, because the culture here doesn't reward and foster platitudes or puffery.

I can get generic advice from many places in the blogosphere, and quite a few people are starting to get it from places like ChatGPT. It's not usually bad advice, but it's mostly simplistic and overview-y.

If a person wants to succeed with this kind of effort, achieve their weight and fitness goals, learning and personalizing are going to be key, IMO. If a person wants to find community and gain acceptance or respect in a community like MFP, then I think sincerity and honesty about where they are on the path is going to outweigh generic posts every time. Early on, that means asking honest questions that might reveal gaps in knowledge or practice, and that's OK . . . perfect, actually.

Just my opinions throughout, though, as usual.

I do agree that that post is suspiciously, "answer my question-y."

I wonder if MFP is trying out auto-generated replies.

Forums. What could possibly go wrong?

I meant the post from that ID on this thread wasn't really bad advice. I agree that the AI sites often provide bad advice, and pretty much always lack any helpful nuance even if not completely inaccurate.

I don't know why anyone trusts them with anything important. Basically, they're spitting out stuff that's the average of what the total internet says (yikes) but will also hallucinate "facts" at times.

ETA: Grammarly offers various tools, not just grammar checking. One tool analyzes text to guess whether it was written by an AI. I'm sure that analyzer is also an AI. 🤣